As a data scientist, I have a keen interest in sentimental analysis since I
will apply text mining and sentiment analysis techniques on the job.
Our agency receives complaints daily concerning issues related to the
United States Postal Service's ability to meet its Universal Service
Obligation (USO). The USO is set up to ensure that Americans receive
mail quickly and pay reasonable (but not exorbitant) costs for mail and
package delivery (Report on Universal Service and the Postal
Monopoly, n.d.). While it is true that the Postal service has a monopoly
on specific services, it cannot indiscriminately raise prices to cover
operational costs. There is a price cap on how much it can charge,
which benefits our country since nearly everyone relies on mail
services.
However, Postal struggles with paying for all the equipment, materials,
and personnel it needs to meet the USO due to the price cap. So what
does this have to do with sentiment analysis? Our agency has been
researching and recommending ways for Postal to remain competitive
while not exceeding the price cap. Our agency completed one body of
work in March 2016, where we outlined several USO funding options
(Funding the Universal Service Obligation, 2016). One option was to
give the Postal service more flexibility in pricing products and services
for which it had a monopoly (Funding the Universal Service Obligation,
2016). Another option included diversifying the products and services
offered by Postal, which might include non-mail services (Funding the
Universal Service Obligation, 2016). These non-mail services might
include driver and fishing license issuance, currency exchange and
banking services, US passport, citizenship, naturalization services, and
other services not traditionally associated with the United States
Postal Service.
If Postal implemented such services, it would need to measure
customer sentiment based on customer feedback on its website or at
post offices with retail point-of-sale terminals. One could then
leverage customer sentiment to drive decisions and measure how
popular or unpopular a new service was to customers.
This discussion brings me to another point. As a data scientist, one
does not need to limit oneself to just using text from Twitter or
Facebook feeds to analyze sentiment. The textual data could reside in
a SQL or NoSQL database populated using a customer survey form
coupled with Postal's Informed Delivery service data, which notifies
customers when mail is delivered, or a product/service is purchased.
Imagine if one could capture subtle data like facial expressions using
an Edge AI solution at the point of sale and then use that information
coupled with the results of sentimental analysis to measure customer
satisfaction.
Another area of interest for me is sentiment analysis of bot farms and
conspiracy theorists seeking to propagate misinformation and sow
discord in our great country. This issue was all over the news during
the 2016 and 2020 elections, so I will say that this area interests me
without getting political. One could use sentiment analysis from social
media posts to analyze the posts' effect and impact on both policy and
public opinion. What posts generate the strongest and weakest
sentiment? What posts generate the most retweets? How many new
followers does an influencer gain following a post? How are social
media posts weaved into the fabric of our society, and how might
those social media posts tear us apart as a country? Where do the
posts originate from, based on geolocation metadata? All important
questions to consider as a data scientist learning to wield the power of
text mining and sentimental analysis.
I am looking forward to learning more about this exciting area of data
analytics.
References:
"Report on Universal Service and the Postal Monopoly" (n.d.).
USPS
.
Retrieved from https://about.usps.com/what/strategic-plans/postal-
act-2006/universal-postal-service.htm
"Funding the Universal Service Obligation" (2016, March 21). United
States Postal Service Office of Inspector General. Retrieved from
https://www.uspsoig.gov/document/funding-universal-service-
obligation/
